Redfish
Diamond Member
Many people think that in a poll of 500 people, the pollster contacts only 500 people and that's the poll, not so. Typically, in scientific, the pollsters will contact far more people than in the final sample because many of those contacted will not be counted in order for the poll to be represent of the population. If the criteria used to construct the sample is party favored, age, family income, race, sex then those selected will form a statically average that matches the area being polled.So they are used for things but not accurate. They are extremely important but they are only accurate in certain circumstances. Companies and political candidates pay big money for polls that are not accurate or are only accurate in certain circumstances or certain instances.Polls don't tell us whether a car, a beer, or politician candidate is good or bad. They only tell us what people think.Ah sarcasm eludes you. If polls were able to predict so well there would only need to be one car built, one beer offered for sale, no need for people to vote. The problem is science wants to believe it understands how humans think and act, therefore it can predict everything about them.Polling services do exactly that. They tell manufactures the kind of product the consumer wants, marketing the kind of advertising most likely to produce results, and policy makers what the public wants from government. However, polls are not the only forecasting tools used by businesses and government.So we should be able to determine everything by polling then. It will make shopping so much easier as we can poll everyone to see what they want to buy, what car to sell, what dishwashing soap to stock, what beer to sell. Imagine just how handy it will be to know in advance who the governor will be, who will be police chief.
The problem is it is very far from even coming close.
Polls are only accurate when people have strong opinions or preferences. For example if a poll was done to determine the best table sale, it would be worthless because there is little difference in taste and most people don't have any opinion. Politician polls are much more accurate because most people have strong opinions and preferences in regard to candidates and issues. Studies have shown that political polls are among the most accurate, about 80%. Polls on the death penalty vary greatly because people vacillate. I have never seen a poll on beer, however you can bet that people who market and advertise the products have.
In politics, government, and business polling is extremely important because it tells us what people think about issues, political candidates, consumer products, and public policy. Governments, politician candidates, and business spend billions of dollars on research studies and polls to determine public opinion. If it was worthless, they would not be spending that kind of money on it.
Got it.
So if polls are only accurate in certain instances and certain circumstances how accurate are they if you have say an extra 10% of the people polled that lean Democrat? What happens to the extreme accuracy of a poll that has a slightly larger number of undecided? What happens to this supposed accuracy if someone feels a need to keep from being chastised by a pollster?
If a poll reveals a much larger number of undecided voters than expected, then the confidence level of poll will be low. How polls handle undecided voters varies by polls. Some polls eliminated them because they believe that they will split in the same manner as decided voters. However most polls report them as undecided.
Keep in mind, pollster don't call elections. They just report how the population being polled plans to vote or who they favor. It is forecasters in news organization that use polling data and other research to call elections.
Oh geez, another nice try, but inaccurate. when the pollsters select a sample of 500 or 1000 they attempt to include a proportional mix of every demographic in the population being sampled. In a diverse country of 330,000,000 its simply impossible to proportionally include every demographic in a sample of 1000.
and you refuse to accept the fact that the pollsters are paid by someone with an agenda. like it or not, today's polls are propaganda, not polling.